Vehicle access information management method and system

Abstract

The application relates to the field of new-generation information technology, and discloses a vehicle access information management method and system to improve performance. The method comprises the following steps: an edge server creates an A-gate group and sets a unified safety interval threshold between any two exit gates in the group, and then establishes a jurisdiction relationship between the A-gate group and at least two exit gates of a merging section; wherein, the gate group attached to any exit gate is unique, and adjacent exit gates with a line distance less than the safety threshold and with a line-of-sight obstacle are included in the same gate group; after receiving a gate opening request of a gate a1 facing a closed state in the A-gate group, it is judged whether there is a conflicting lateral vehicle on the lateral exit of other gates in the group, and if there is, a delay waiting process is performed on the exit gate opened later to make the time interval between the gate opening instruction of the exit gate opened later and the last gate closing instruction of other exit gates in the gate group greater than or equal to the safety interval threshold.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of new generation information technology, and in particular to a vehicle access information management method and system. BACKGROUND

[0002] Vehicle access information management has been widely used in schools, communities, shopping malls and other places.

[0003] The existing vehicle access information management system mostly exists in the situation of losing one side in order to gain the other in terms of cost and performance. On the one hand, in the use scene with high performance requirements, the system needs high-configuration server clusters and cloud storage spaces to store and coordinate data for the perfection of functions. On the other hand, in most ordinary use scenes, the property party deploys relatively low-configuration software and hardware management resources based on cost considerations, and the low-configuration resources mostly have management loopholes and are prone to generate bad behaviors such as escaping orders. Occasionally generated bad behaviors are usually regarded as small probability events in the past, but in the current situation of rapid spread of short videos, bad behaviors are easy to spread and cause more imitations, thereby causing adverse effects on property management and social atmosphere.

[0004] At the same time, even if the high-configuration vehicle access information management system mostly provides products with standard configuration and fixed functions, which are separated from the environmental information in the deployment scene, there are subsequent security risks in some use scenes: when two independent vehicle accesses are close to each other, on the one hand, some drivers of tourist identity have no prior knowledge of the road conditions; on the other hand, even the drivers familiar with the road conditions are prone to have a decrease in prediction ability; the inherent building hinders the driver's line of sight, especially in the environment where some garages are loaded with roller garage doors, which is easy to cause collision accidents of vehicles at the two close accesses during the merging process; and the probability and severity of the accidents are much higher than those at the intersection of the internal roads of the garage, because the drivers can mostly determine in advance and slow down and use the warning light at the intersection of the internal roads of the garage with the resources such as ceiling lamps and wide-angle mirrors.

[0005] Therefore, based on the common hidden dangers of the high and low configurations of the vehicle access software and hardware management resources in different use scenes and the separation from the environmental information, it is necessary to develop related technologies for patching, plugging and other ways to repair and flexibly match the existing vehicle access information resources with the environmental information at a low cost. SUMMARY

[0006] The present application aims to disclose a vehicle access information management method and system to improve performance.

[0007] To achieve the above purpose, the vehicle access information management method disclosed by the present application comprises:

[0008] Step S1, the edge server creates an A gate group according to the artificial request and sets a unified safety interval threshold Δt between any two exit gates in the group, and then establishes the jurisdiction relationship between the A gate group and at least two exit gates of the merging and converging road section according to the artificial input information; wherein, the gate group attached to any exit gate is unique, and the adjacent exit gates with a line distance less than the safety threshold and existing visual obstacles are included in the same gate group;

[0009] Step S2, after receiving the opening request of the gate a1 facing the closed state in the A gate group, the edge server records the time stamp of the opening request, and then judges whether the difference between the current time t1 and the last closing time t0 of other gates in the group is less than the safety interval threshold; if yes, wait until t0+Δt and then go to step S3; if no, go to step S3 directly;

[0010] Step S3, the edge server judges whether there is a waiting lateral vehicle in the current A gate group, if yes, determines the opening sequence between the conflicting exit gates in the group according to the time stamp of the recorded opening request in sequence, and then executes a delay waiting process on the later opened exit gate when issuing the opening instruction to the earlier opened exit gate, so that the time interval between the opening instruction issued to the later opened exit gate and the last closing instruction of other exit gates in the gate group is greater than or equal to the safety interval threshold.

[0011] Preferably, the edge server judging whether there is a conflicting lateral vehicle in the gate a2 in the A gate group includes one or both of the following two judgment dimensions:

[0012] Dimension one, a camera-equipped license plate recognition instrument is used to monitor whether there is a motor vehicle of the merging and converging road section at the lateral exit where the gate a2 is located; the exit industrial computer is in communication connection with the license plate recognition instrument and the edge server, and the opening request of the motor vehicle is sent by the exit industrial computer after verifying that the motor vehicle meets the release conditions;

[0013] Dimension two, a card reader deployed at the lateral exit where the gate a2 is located is used to monitor whether there is a non-motor vehicle of the merging and converging road section; the non-motor vehicle communicates with the card reader through radio frequency communication by means of a magnetic identification card; wherein, after identifying the tag information of the magnetic identification card, the card reader sends an opening request carrying a time stamp to the edge server.

[0014] Preferably, when dimension one and dimension two exist at the same time, the method further comprises:

[0015] If the edge server receives an opening gate request sent by the card reader or the exit industrial computer in the lateral exit where the gate a2 is located after issuing the opening gate instruction to the gate a2 and before issuing the closing gate instruction to the gate a2 according to the timer, the edge server resets the timer that maintains the open state of the gate a2 to restart timing after the first timing length of the countdown, and then issues the closing gate instruction to the gate a2 after the timing ends.

[0016] If the edge server receives at least two opening gate requests composed of dimension one and dimension two in the lateral exit where the gate a2 is located before issuing the opening gate instruction to the gate a2, the edge server combines the at least two opening gate requests into the same opening gate request and saves the timestamp according to the earliest timestamp in the combined opening gate request, and sets the timing length of the timer to twice the first timing length when issuing the opening gate instruction to the gate a2.

[0017] Preferably, the method further comprises:

[0018] If the edge server receives opening gate requests composed of at least two dimension twos in the lateral exit where the gate a2 is located before issuing the opening gate instruction to the gate a2, the edge server combines the at least two opening gate requests into the same opening gate request, and sets the timing length of the timer to the first timing length when issuing the opening gate instruction to the gate a2.

[0019] Preferably, the application further comprises:

[0020] The interactive terminal of the exit where the gate is located after the indication of the edge server sends notification information of delay waiting through the display screen and / or the voice module to inform the user of the reason for delay waiting before the gate is opened.

[0021] Preferably, the application further comprises:

[0022] The edge server collects video images of the motor vehicle after the exit gate is opened through the camera of the license plate recognition instrument, and tracks the displacement trajectory of the extracted license plate relative to the camera in the video images, and when the motor vehicle driving direction is identified as reverse according to the displacement trajectory, the file name of the video images is combined with the collection time and the identified license plate number to rename and save to an abnormal image library for evaluating whether the intention is to escape single, in response to the retrieval request of the user based on the time period and / or license plate information.

[0023] Preferably, the application further comprises:

[0024] The edge server extracts at least two key frames from the video image in the driving-off direction, then performs a body consistency comparison between the key frames in sequence and the image frames collected by the entrance barrier for confirming the license plate storage time of the same motor vehicle, when the consistency comparison result of any key frame is greater than or equal to a set consistency threshold, the subsequent comparison is terminated and the video image is discarded; if the comparison results of all key frames are less than the consistency threshold, the video image is renamed by combining the file name with the collection time and the identified license plate number, and then saved to a second type of abnormal image library for evaluating whether there is a potential traffic accident escape in the garage, to respond to the retrieval request of the user based on the time period and / or license plate information.

[0025] Preferably, the body consistency comparison algorithm specifically includes:

[0026] Identifying and separating the vehicle body from the background;

[0027] Anchoring a group of reference frame points and a group of high-collision-feature points from the vehicle body;

[0028] According to the positional relationship of the reference frame points, the geometric transformation relationship of the reference frame points in the key frame mapped to the positional reference frame points in the storage collection image frame is estimated, and then the vehicle body image in the key frame is resampled according to the geometric transformation relationship to realize the posture and scale alignment of the vehicle body image in the key frame with the vehicle body image in the storage collection image frame;

[0029] The gray value jump value sequence of the line segment connected between the two high-collision-feature points is obtained;

[0030] The approximation degree of the gray value jump value sequence between the aligned vehicle body images of the two image frames is compared one by one, and if the approximation degree of the gray value jump value sequence of any positional line segment is less than the set consistency threshold, it is determined that the consistency comparison result of the current key frame is less than the set consistency threshold.

[0031] Preferably, the reference frame points include: roof corner points, window frame corner points and hub center recesses; and the high-collision-feature points include: lamp group corner points of the front face, edge corner points of the bumper, bottom end points of the side door beltline, and front end points of the door handle.

[0032] In order to achieve the above purpose, the application further discloses a vehicle entry and exit information management system, which comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor implements the above method when executing the computer program.

[0033] The application has the following beneficial effects:

[0034] 1. The existing vehicle access information management platform mostly adopts cloud-edge-end architecture, the application mainly adds and optimizes the control level function in the edge server, the change of the existing system is small, new hardware resources are not needed to be introduced additionally, and the added function can be conveniently and widely applied to the upgrading and reconstruction of the existing vehicle access information management resources.

[0035] 2. The core of the application is that the edge server creates an A barrier group and sets a unified safety interval threshold between any two exit barriers in the group, and establishes a jurisdiction relationship between the A barrier group and at least two exit barriers of the merging road section; after receiving the opening request of the barrier a1 facing the closed barrier in the A barrier group, it is judged whether there is a conflicting lateral vehicle in the reasonable time period before and after the current time in the lateral exit of the other barrier in the group, if there is, the barrier opened after the conflict is executed to delay and wait to make the time interval between the opening instruction of any exit barrier in the barrier group and the last closing instruction of the other exit barrier greater than or equal to the safety interval threshold; thereby fundamentally avoiding the possibility of collision of multiple vehicles in different directions during merging; the overall logic is rigorous and reliable, compared with the existing management system, the overall performance is significantly improved. Moreover, the safety interval threshold can be flexibly differentiated in different environments, which is flexible and easy to implement and promote.

[0036] 3. The application has good expansibility, especially when the motor vehicle and the non-motor vehicle are mixed at the exit, the overall travel efficiency can be significantly improved, for example: after receiving the opening request of the barrier a1 facing the opened barrier in the A barrier group, the current opening request and the last opening request of the barrier a1 can be combined, the timer is reset to ensure that the continuous opening requests facing the barrier a1 in a short time are all processed, and then the opening demand processing of other barriers is switched, the number of switching between different barriers is reduced, and the data processing pressure of the edge server is also reduced.

[0037] The application will be further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0038] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application, and are incorporated in and constitute a part of this application. The embodiments of the application illustrated in the drawings, and their description thereto, are presented to provide the practitioner with a thorough and enabling disclosure of the application, and are included as a part of this application to convey the scope of the application to those skilled in the art. In the drawings:

[0039] Figure 1 is a flowchart of the vehicle access information management method disclosed by the embodiment of the application.

[0040] Figure 2 is a distribution diagram of two adjacent exit barriers in the same barrier group whose distance is less than the safety threshold and there is a line of sight obstacle.

[0041] Figure 3 is a logical flow chart of the consistency detection algorithm disclosed by the embodiments of the present application. DETAILED DESCRIPTION

[0042] The embodiments of the present application are described in detail below with reference to the accompanying drawings, but the present application can be implemented in various different ways limited and covered by the claims.

[0043] Embodiment one

[0044] The present embodiment discloses a vehicle access information management method, as shown in the figure, comprising the following steps: Figure 1

[0045] Step S1, the edge server creates an A barrier group according to the manual request and sets a unified safety interval threshold Δt between any two exit barriers in the group, and then establishes the jurisdiction relationship between the A barrier group and at least two exit barriers of the merging route section according to the manually input information.

[0046] In this step, any exit barrier is attached to a unique barrier group; and with reference to Figure 2 , the present embodiment restricts the adjacent exit barriers with a line distance less than the safety threshold (usually less than twice the length of a family sedan, usually unattended, and if there is one, it is basically responsible for supervision by a single on-duty personnel of the same security booth, and the specific value can be set within the range of 5-10 meters) and there is a line of sight obstacle into the same barrier group. Optionally, the user can arrange one, two or more barrier groups according to the specific environment of the implementation scene; the safety interval threshold of each barrier group can be individually assigned according to the actual situation of the test. Among them, the processing logic of each barrier group is the same, and the present embodiment only takes barrier group A as an example to avoid mixing different barrier groups, and the following will not be described.

[0047] Step S2, after receiving the opening request of barrier a1 facing the closed state in A barrier group, the edge server records the timestamp of the opening request, and then judges whether the difference between the current time t1 and the last closing time t0 of other barriers in the group is less than the safety interval threshold, if yes, wait until t0+Δt and then go to step S3; if not, go to step S3 directly.

[0048] ​In this step, if the opening request of the barrier a1 facing the open state is received, the current opening request of the barrier a1 can be combined with the previous opening request. Referring to the processing mode of the barrier a2 in case one of the subsequent embodiment 2, the timer can be reset to ensure that the continuous opening requests of the barrier a1 within a short time are processed and then the opening request of other barriers is processed. Meanwhile, the "closing time" of this step is the time when the edge server issues the closing instruction, which will not be described in detail.

[0049] In step S3, the edge server determines whether there is a waiting lateral vehicle in other barriers in the A barrier group. If there is, the opening sequence between the conflicting exit barriers in the group is determined according to the time stamp of the recorded opening request, and then the delay waiting processing is performed on the later opened exit barrier when the opening instruction of the first opened exit barrier is issued, so that the time interval between the opening instruction of the later opened exit barrier and the previous closing instruction of the other exit barrier in the barrier group is greater than or equal to the safety interval threshold.

[0050] In this step, on the contrary, if it is determined that there is no waiting lateral vehicle in other barriers in the A barrier group, the edge server directly issues the opening instruction of the barrier a1.

[0051] In this step, based on the fact that other barriers in the group may also have the same waiting opening vehicle as the barrier a1, resource conflict will occur. In this embodiment, the conflict is processed based on the time stamp of each opening request. When the time stamps are inconsistent, the first-in-first-out principle is adopted to make a decision. In a small number of cases, if the time stamps of two or more conflicting opening requests are consistent, a random method can be used to determine the opening sequence between the conflicting exit barriers. In order to accurately and reasonably determine the sequence of each barrier in the group, the edge server configures a queue in the cache for each barrier group, and then maintains the opening request to be processed between different barriers in the group in the queue in a first-in-first-out manner. The specific implementation is a prior art, which will not be described in detail.

[0052] Embodiment two

[0053] This embodiment is based on the above-mentioned embodiments and considers the case of mixed use of motor vehicles and non-motor vehicles in the exit barrier. Non-motor vehicles mainly include motorcycles and three-wheeled vehicles for cleaning.

[0054] In the scenario set in this embodiment, the edge server determines whether there is a conflicting lateral vehicle in the barrier a2 in the A barrier group according to one or both of the following two judgment dimensions:

[0055] Dimension one, the vehicle license plate recognition instrument with a camera monitors whether there is a motor vehicle that converges into the merging section at the lateral exit of the barrier a2; the exit industrial computer establishes a communication connection with the license plate recognition instrument and the edge server, and the opening request of the corresponding motor vehicle is sent by the exit industrial computer after verifying that the motor vehicle meets the release condition.

[0056] Among them, the exit industrial computer is the local brain of the lane (that is, the end in the three-level architecture of cloud-edge-end), which integrates management software for coordinating license plate recognition instruments, light supplementing lamps, voice modules, and display screens to process local logic, including but not limited to opening instructions or closing instructions issued by the edge server to the exit barrier.

[0057] Dimension two, the card reader deployed at the lateral exit of the barrier a2 monitors whether there is a non-motor vehicle that converges into the merging section; the non-motor vehicle communicates with the card reader through radio frequency communication with a magnetic identification card; wherein the card reader sends an opening request carrying a timestamp to the edge server after identifying the tag information of the magnetic identification card.

[0058] Non-motor vehicles cannot be processed based on existing license plate recognition instruments, and there are many situations such as no license plate, large differences in vehicle type from motor vehicles, and impact on intelligent switching reliability, etc., so that in most application scenarios, the card reader and the exit industrial computer are two independent components, and the card reader is usually added by the property according to the needs of the owners after the deployment of the exit industrial computer and the license plate recognition instrument. As a variation, the exit industrial computer and the card reader can be physically integrated, but the processing logic still needs to be distinguished according to the above two dimensions, so this variation still belongs to the protection scope of the present application, and will not be described in detail.

[0059] Further, when dimension one and dimension two exist at the same time, the method further comprises:

[0060] Case one, the edge server, if it receives an opening request sent by the card reader or the exit industrial computer at the lateral exit of the barrier a2 after issuing an opening instruction to the barrier a2, and before issuing a closing instruction to the barrier a2 according to the timer, it will reset the timer that maintains the opening state of the barrier a2 to a first time length of countdown and then restart the timer, and then issue a closing instruction to the barrier a2 after the timer ends. Optionally, for a roller shutter barrier, the first time length can be set in the range of 5-10 seconds based on the configuration of most motors; for a straight rod, the range of the first time length can be reduced to 3-6 seconds.

[0061] Further, in the present embodiment, if the export industrial computer supports analyzing the video collected by the camera of the license plate recognizer to identify whether the vehicle has completed passing, etc. (which can usually be achieved by extracting key frames of the video at an interval period and comparing them with a pre-labeled static image of no-vehicle passing for consistency), the information confirming that the vehicle has passed can be fed back to the edge server, so that the edge server can reset the timer to zero and issue a gate closing instruction before the timing of the timer ends.

[0062] Case two, if the edge server receives at least two opening gate requests composed of dimensions one and two in the lateral export on the side of the barrier a2 before issuing an opening gate instruction to the barrier a2, the at least two opening gate requests are merged into the same opening gate request, and the timestamp is saved with the earliest timestamp in the merged opening gate request. When the opening gate instruction is issued to the barrier a2, the timing length of the timer is set to twice the first length.

[0063] In the above case two, due to the field of view angle limitation of the license plate recognizer camera, the number of motor vehicles is only one in actual situations, so the length of the timer should not be too large; at the same time, in order to effectively avoid the rear-end collision or scratching of motor vehicles and non-motor vehicles due to congestion or sprinting at the same export barrier, the length of the timer should not be consistent with case one; therefore, doubling the length of the timer countdown time is a more reasonable strategy that conforms to the actual situation; moreover, even if the number of non-motor vehicles is relatively large, causing the length of case two in the execution process to be insufficient for the safe passing of the following vehicle, the system can automatically switch to case one to further extend the actual length of the export barrier in a single release; that is, the two cases can not only complement each other but also perform emergency switching, thereby unexpectedly improving the flexible compatibility of the actual use scene and the safety performance of the system as a whole.

[0064] Preferably, the method further comprises:

[0065] Case three, if the edge server receives opening gate requests composed of at least two dimensions two in the lateral export on the side of the barrier a2 before issuing an opening gate instruction to the barrier a2, the at least two opening gate requests (without opening gate requests of dimension one) are merged into the same opening gate request, and when the opening gate instruction is issued to the barrier a2, the timing length of the timer is set to the first length (wherein the objects and meanings of the timer and the first length are consistent with the above two cases, and are not described again).

[0066] Obviously, the above three cases described in the embodiment can significantly improve the travel efficiency when the motor vehicle and the non-motor vehicle use the exit gate together. Further, to improve the user experience, the method can be further supplemented and optimized, for example: the interactive terminal of the exit gate indicated by the edge server sends notification information of delayed waiting through the display screen and / or voice module before opening the gate to inform the user of the reason for the delayed waiting.

[0067] Embodiment three

[0068] The embodiment is further functionally extended on the basis of any of the above embodiments.

[0069] In the embodiment, the edge server collects video images of the motor vehicle after the exit gate is opened by the camera of the license plate recognition instrument, and tracks the displacement trajectory of the extracted license plate relative to the camera from the video images. When it is identified according to the displacement trajectory that the motor vehicle is driving in the reverse direction, the file name of the video image is combined with the collection time and the identified license plate number to rename and save to an abnormal image library for evaluating whether it is intended to escape the single, in response to the user's retrieval request based on the time period and / or license plate information.

[0070] In contrast, most of the existing garage management systems do not save the images or videos of the vehicle leaving the garage due to limited storage resources. The image of entering the garage is usually only a single photo of extracting the license plate information for determining the starting time of charging. At the same time, the existing garage management system considers that the motor vehicle with the current collected license plate has been processed as completing the exit after the exit gate is opened. When the motor vehicle is driving in the reverse direction, it is considered as entering the garage or classified as an abnormality that can be ignored and directly released, resulting in escaping the single.

[0071] In the embodiment, the video image after the gate is opened can be collected for a preset time length. When leaving the garage normally, the displacement trajectory of the license plate extracted from multiple key frames relative to the camera presents a gradual change rule from far to near and then disappears (the rule is also the translation rule of the center coordinate point of the license plate in the same coordinate system of each key frame alignment); and if the displacement trajectory of the license plate extracted from multiple key frames relative to the camera presents a gradual change rule from far to near, and then suddenly changes from near to far until disappears, the driving direction of the motor vehicle can be identified as reverse according to the displacement trajectory.

[0072] Based on the above processing of the embodiment, on the one hand, it avoids massive storage, and on the other hand, it can be used as evidence for the manager to intervene in advance, and the specific intervention methods include but are not limited to correcting the exit records that are contrary to the facts, or marking the related vehicle as an abnormal vehicle to limit its re-leaving the garage to be disposed by the on-duty personnel.

[0073] Preferably, the collection time of the composition renaming is arranged before the license plate, facilitating the time sequence of such video when stored, and further facilitating the on-duty personnel to intervene in sequence. At the same time, the video is renamed by the collection time, which can effectively avoid the interference caused by attribute changes in the subsequent processing process such as moving or forwarding.

[0074] Further, the method of the embodiment further comprises:

[0075] The edge server extracts at least two key frames from the video image in the driving direction from the exit direction, and then performs consistency comparison of the vehicle body on the key frames in sequence and the image frames collected by the entrance barrier for confirming the license plate storage time of the same motor vehicle. When the consistency comparison result of any key frame is greater than or equal to a set consistency threshold, the subsequent comparison is terminated and the video image is discarded; if the comparison results of all key frames are less than the consistency threshold, the video image is renamed by combining the collection time and the identified license plate number, and then saved to a second type of abnormal image library for evaluating whether there is a potential traffic accident escape in the garage, to respond to the retrieval request of the user based on the time period and / or license plate information.

[0076] Similarly, as a comparison, the existing garage management system has no similar function, and the purpose of consistency verification involved in a small amount of existing technology is usually limited to judging whether it is a fake license plate vehicle; therefore, it is irrelevant to the main purpose of the embodiment for evaluating whether there is a potential traffic accident behavior in the garage.

[0077] Preferably, as shown in Figure 3 The consistency comparison algorithm of the vehicle body specifically includes:

[0078] Step S51, identifying and separating the vehicle body from the background.

[0079] Step S52, anchoring a group of reference frame points and a group of high-collision feature points from the vehicle body.

[0080] The so-called "reference frame points" and "high-collision feature points" both belong to a group of pre-defined key positions that are stable on most vehicles and easy to identify in images; the "reference frame points" are usually reference points that are not easily damaged, and the so-called "high-collision feature points" refer to reference points whose connection region with similar points is easily damaged.

[0081] Optionally, the reference frame points include but are not limited to roof corner points, window frame corner points, and hub center recesses; the high-collision feature points include but are not limited to the lamp group corner points of the front face, the edge corner points of the bumper, the bottom end points of the side door waist line, and the front end points of the door handle.

[0082] Step S53, according to the alignment relationship of the reference frame points, the geometric transformation relationship of the reference frame points in the key frame mapping to the alignment reference frame points in the warehouse collection image frame is estimated, and then the vehicle body image in the key frame is resampled according to the geometric transformation relationship to realize the relative alignment of the vehicle body image in the key frame and the vehicle body image in the warehouse collection image frame in the posture and scale.

[0083] Step S54, the gray level jump value sequence of the line segment connected between the two collision-prone feature points in the adjacent pixels is formed.

[0084] Step S55, the approximation degree of the gray level jump value sequence between the alignment vehicle body images in the two image frames is compared one by one, if the approximation degree of the gray level jump value sequence of any alignment line segment is less than the set consistency threshold, it is determined that the consistency comparison result of the current key frame is less than the set consistency threshold. On the contrary, if the approximation degree of the gray level jump value sequence of all alignment line segments is greater than or equal to the set consistency threshold, it is determined that the consistency comparison result of the current key frame is greater than or equal to the set consistency threshold.

[0085] Optionally, the specific calculation index of the approximation degree can adopt the Pearson correlation coefficient, and the corresponding consistency threshold can be calibrated according to the empirical value.

[0086] Based on the above consistency detection algorithm, the processing process focuses on the alignment of the collision-prone feature points and the approximation degree calculation between the connected lines, which can ensure the accuracy and real-time performance of the processing result.

[0087] Based on the above processing of the embodiment for evaluating whether the traffic accident behavior in the potential garage occurs, the suspicious vehicle and its credentials for the on-duty personnel to focus on can be automatically screened out. When the conditions allow, the on-duty personnel can call the video data recorded by other cameras in the garage to trace back the running track of the suspicious vehicle during the verification process. Therefore, the embodiment can improve the work efficiency of the on-duty personnel, thereby reducing the configuration cost of the on-duty personnel and the video supporting storage resources, and also extends the utilization value of the stored video.

[0088] Embodiment four

[0089] The embodiment discloses a vehicle access information management system, comprising a storage, a processor, and a computer program stored in the storage and executable on the processor, and the processor implements the method corresponding to the above-mentioned embodiments one to three when executing the computer program.

[0090] In summary, the vehicle access information management method and system disclosed in the above four embodiments of the application have at least the following beneficial effects:

[0091] 1. The existing vehicle access information management platform mostly adopts cloud-edge-end architecture, the present application mainly adds and optimizes the control level function in the edge server, the change of the existing system is small, without introducing new hardware resources, the added function can be conveniently and widely applied to the upgrading of the existing vehicle access information management resources.

[0092] 2. The core of the present application is that the edge server creates an A barrier group and sets a unified safety interval threshold between any two exit barriers in the group, and then establishes the jurisdiction relationship between the A barrier group and at least two exit barriers of the merging road section; after receiving the opening request of barrier a1 facing the closed state in the A barrier group, it is judged whether there is a conflicting lateral vehicle in the current and previous reasonable time period of the lateral exit of other barriers in the group, if there is, the delay waiting processing is performed on the barrier which is opened later to cause the time interval between the opening instruction of any exit barrier in the barrier group and the last closing instruction of other exit barriers to be greater than or equal to the safety interval threshold; thereby fundamentally avoiding the possibility of collision of multiple vehicles in different directions during merging; the overall logic is rigorous and reliable, compared with the existing management system, the overall performance is significantly improved. Moreover, the safety interval threshold can be flexibly differentiated in different environments, which is flexible and easy to implement and promote.

[0093] As a comparison, if the safety interval is set as the interval between the two opening instructions of different barriers in the barrier group, it cannot be compatible with the above-mentioned merging processing of multiple opening requests when non-motor vehicles are mixed in the lane, thereby affecting the traffic efficiency due to the frequent switching between different barriers. On the contrary, if the merging processing of multiple opening requests is still performed in such a scene, there is a loophole that the closing time of the first opening exit barrier cannot be estimated, thereby easily causing the time interval between the vehicles merging from different directions in the barrier group to be actually less than the safety interval threshold, thereby leading to an unavoidable collision accident. Therefore, the present application restricts the safety interval threshold to the safety interval between the last closing instruction of one exit barrier and the next opening instruction of the adjacent exit barrier in the group, ensuring that the overall logic is more rigorous and reliable.

[0094] 3. The present application has good expansibility, especially when motor vehicles and non-motor vehicles are mixed in the exit, it can significantly improve the overall travel efficiency, for example: after receiving the opening request of barrier a1 facing the opened state in the A barrier group, the current opening request and the last opening request of the barrier a1 can be merged, by resetting the timer, to ensure that the continuous opening requests facing the barrier a1 in a short time are all processed, and then switch to the opening demand processing of other barriers, effectively reducing the switching times between different barriers, and also reducing the data processing pressure of the edge server.

[0095] The above merely provides the preferred embodiments of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the principles and technical scope of the present application shall fall into the scope of the present application.

Claims

1. A method for information-based management of vehicle access, characterized in that, include: Step S1: The edge server creates a gate group A based on manual requests and sets a uniform safety interval threshold Δt between any two exit gates in the group. Then, based on the information input by the manual server, it establishes the jurisdiction relationship between gate group A and at least two exit gates that merge into the confluence section. Among them, any exit gate is attached to a unique gate group, and adjacent exit gates with constraint line distance less than the safety threshold and visual obstruction are included in the same gate group. Step S2: After receiving the opening request from the A-barrier group for the closed gate a1, the edge server records the timestamp of the opening request, and then determines whether the difference between the current time t1 and the most recent closing time t0 of other gates in the group is less than the safety interval threshold. If so, it waits until t0+Δt before proceeding to step S3; otherwise, it proceeds directly to step S3. Step S3: The edge server determines whether there are any waiting lateral vehicles in other gates in the current A gate group. If so, it determines the opening order of the conflicting exit gates in the group according to the timestamps of the recorded gate opening requests. Then, when issuing the gate opening command to the exit gate that opens first, it performs a delay waiting process on the exit gate that opens later so that the time interval between the gate opening command issued to the exit gate that opens later and the previous gate closing command of other exit gates in the gate group is greater than or equal to the safety interval threshold. The edge server determines whether there is a conflicting lateral vehicle approaching from barrier a2 in barrier group A by including one or both of the following two dimensions: Dimension 1: A license plate recognition device with a camera monitors whether there are motor vehicles merging into the merging section at the lateral exit of the barrier gate a2; the exit industrial control computer establishes a communication connection with the license plate recognition device and the edge server, and the gate opening request corresponding to the motor vehicle is issued by the exit industrial control computer after verifying that the motor vehicle meets the release conditions; Dimension 2: The card reader deployed at the side exit where the gate a2 is located monitors whether there are non-motorized vehicles merging into the merging section; the non-motorized vehicles communicate with the card reader via radio frequency through magnetic identification cards; wherein, after the card reader recognizes the tag information of the magnetic identification card, it sends a gate opening request carrying a timestamp to the edge server. When both dimension one and dimension two exist, the method further includes: If the edge server receives an opening request from the card reader or the exit control computer at the side exit where the barrier gate a2 is located after issuing an opening command to the barrier gate a2 and before issuing a closing command to the barrier gate a2 according to the timer, it will reset the timer that maintains the open state of the barrier gate a2 to the first countdown duration and start counting again, and then issue a closing command to the barrier gate a2 after the timer expires. If, before issuing an opening command to the barrier gate a2, the edge server receives at least two opening requests consisting of dimension one and dimension two from the lateral exit where the barrier gate a2 is located, the at least two opening requests are merged into the same opening request and the timestamp saved is based on the earliest timestamp of the merged opening request. When issuing an opening command to the barrier gate a2, the timer duration is simultaneously set to twice the first duration. If, before issuing an opening command to the barrier gate a2, the edge server receives an opening request from the lateral exit where the barrier gate a2 is located consisting of only two dimensions two, the server merges the two opening requests into a single opening request and simultaneously sets the timer duration to the first duration when issuing the opening command to the barrier gate a2.

2. The vehicle access information management method according to claim 1, characterized in that, Also includes: The edge server instructs the interactive terminal at the exit where the gate is located to send a delay notification message via a display screen and / or voice module before the gate is opened to inform the user of the reason for the delay.

3. The vehicle access information management method according to claim 1 or 2, characterized in that, Also includes: The edge server uses the camera of the license plate recognition device to capture video images of motor vehicles after the exit gate is opened, and tracks the displacement trajectory of the license plate relative to the camera extracted from the video image. When the displacement trajectory indicates that the motor vehicle is driving in the wrong direction, the file name of the video image is renamed by combining the acquisition time and the identified license plate number, and then saved to an abnormal image library used to assess whether there is an intention to evade payment, in response to the user's retrieval request based on time period and / or license plate information.

4. The vehicle access information management method according to claim 3, characterized in that, Also includes: The edge server extracts at least two key frames from the video image traveling in the direction of departure. Then, it compares each key frame sequentially with image frames of the same vehicle captured at the entrance gate to confirm the license plate entry time. If the consistency comparison result of any key frame is greater than or equal to a set consistency threshold, the subsequent comparison is terminated and the video image is discarded. If the comparison results of all key frames are less than the consistency threshold, the file name of the video image is renamed by combining the capture time and the identified license plate number, and then saved to a Class II abnormal image library for assessing whether there is a potential hit-and-run incident in the garage, in response to user retrieval requests based on time period and / or license plate information.

5. The vehicle access information management method according to claim 4, characterized in that, The vehicle body consistency comparison algorithm specifically includes: Identify and separate the vehicle body from the background; Anchor a set of reference frame points and a set of high-risk collision feature points from the vehicle body; Based on the alignment relationship of the reference frame points, estimate the geometric transformation relationship between the reference frame points in the key frame and the corresponding reference frame points in the inbound image frame. Then, based on the geometric transformation relationship, resample the entire vehicle image in the key frame to achieve a state of relative alignment in attitude and scale with the entire vehicle image in the inbound image frame. The grayscale jump values ​​of the line segment connecting two high-collision feature points are formed into a sequence between adjacent pixels; The similarity of the sequence formed by the gray-scale jump values ​​of the vehicle body images aligned between two image frames is compared one by one. If the similarity of the sequence formed by the gray-scale jump values ​​of any aligned line segment is less than the set consistency threshold, then the consistency comparison result of the current key frame is determined to be less than the set consistency threshold.

6. The vehicle access information management method according to claim 5, characterized in that, The reference frame points include: the roof corner, the window frame corner, and the center recess of the wheel hub; the high-collision feature points include: the headlight corner, the edge corner of the bumper, the bottom end of the side door waistline, and the front end of the door handle.

7. A vehicle access information management system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method described in any one of claims 1 to 6.

Images